Thanks Roger, I grasped the pblm in the end. What a dolt I am

It is a subspecies of carlinae.
Chris
Interesting. Let me note that of the cited references, neither is a scientific publication, and neither provides any evidence or justification for their opinion (which is generally required in scientific work). The latest comprehensive revision of the Palearctic species of the genus Pyrgus, as far as I know, is the one by R. de Jong, a highly reputed expert on Hesperiidae (Tijdschrift voor entomologie 115, 1-128, 1972). I haven't done a comprehensive search of the literature, so there may be further works on the topic, although probably none based on DNA as there are no publicly available DNA sequences of either species in the Barcode of Life Database. Anyway, the following quote from de Jong's work provides some clues into why different sources choose to treat the subject differently. It goes on discussing larval development differences etc; for those interested, you can find the whole work on the Biodiversity Heritage Library website.Roger Gibbons wrote:My opinion is that cirsii is not a subspecies of carlinae, as evidenced by publications by Lafranchis and the PACA Atlas to name but a few
15. Superspecies Pyrgus carlinae Rambur
Taxonomy. — Until Picard's publication in 1950, carlinae and cirsii were considered separate species, as they differed constantly in facies and genitalia. Picard (1950b) found in the French Alps intermediate specimens almost as numerous as typical specimens, particularly in La Bessée (Hautes Alpes), but he did not give exact data. [....] Apparently, Picard estimated the number of intermediates too high. Guillaumin (1964) mentions a score of 14 % intermediates in the localities where carlinae and cirsii come into contact (Hautes Alpes: La Bessée, Le Lautaret). Additional studies are still in progress (Guillaumin, in litt.). For the present study the most important fact is the interbreeding, as it indicates an incomplete stage of speciation. Contact between the lowland form cirsii and the mountain form carlinae is only known from the French Alps. Apparently, they interbreed wherever they have contact. It is nonsensical to consider them separate species. I prefer to call carlinae and cirsii semispecies, as members of the same superspecies, thus indicating the supposed, rather progressed stage of speciation. But of course, it is also tenable to call them subspecies of a single species, see also Chapter l.b. The decision is not important for the present study.
(De Jong, 1972)